Immigration to Atlanta: Historical Data Visualization

This week before Spring Break, I was fortunate enough to get put on a new project coming out of the History Department. Working with Dr. Marni Davis, a team of SIFs and I are helping Marni to organize, visualize, and present her data on immigrants to Atlanta. We are beginning with data in the late 1800’s, and working up to the present. Currently, we have a lot of data up to about 1930, which is what I have been working with.

To begin, Marni supplied us with a spreadsheet of data with about 1600 entries on immigrants. These entries have data points such as name, birthdate, port entry city, date of immigration into Atlanta, date of naturalization, country of origin, address in Atlanta when applying for citizenship, and so on. Because I have experience with Tableau (a program which can produce beautiful data visualizations), Marni asked me to take some of this data and create charts that we could put up on her new GSU sites website dedicated to the Immigrants ATL project.

To begin, I decided to try and create a bit of a story that showed very simple data. For example, in the first figure, I show the m/f immigration difference, where you can see that men immigrated to Atlanta 100x more than women in this period.

Immigrant-Gender-1

In the next figure, you can see a comparison between when Asian/Pacific peoples were immigrating into Atlanta, versus when Central Europeans were. Additionally, you can see that there were far more Central Europeans immigrating to Atlanta than there were Asians.

Region Immigration Comparison

The visualizations are really wonderful in that they make the data much easier to mentally process and compare, and they will be easy to present in any venue. Further, we can arrange any data visualization into a kind of story that we want the data to tell.

Currently we are working to create more graphs like you see above, but also to incorporate some maps that show data such as average age of immigrants into Atlanta from various regions, countries, and cities.

While these visualizations are gorgeous, and not difficult to make, there are some issues that arise to complicate matters. For example, there is not a year of entry for absolutely every immigrant to Atlanta. In order to create the graphs, I have to omit whole people who may only be missing one piece of data. Further, I learned the hard way, that Tableau does not read the formulas that we make in Excel. I had a lot of loading issues and eventually found that Tableau is set up to do my computing. After several hours of trying to load, visiting the Tableau sub-reddit, and doing a lot of Googling, that I could subtract the Immigration year from the Naturalization year right in the graph.

Next, we are going to build out more visualizations, meet with the rest of the team to see what they are working on, and hopefully create a really robust and face-smackingly wonderful set of data that Marni can present in any venue easily as she works to collect and manage all this data.

The Tableau Experience

Today I attended “The Tableau Experience.” I’ve been struggling with what to call it: a conference, a convention… It had only one speaker, and a Q&A – and a wet-bar.  So… I’m settling with ‘experience’ as they do. What it really was though, was an advertisement.

Don’t get me wrong – I like using Tableau to recreate charts for the Tobacco Ebook I’m working on, and this ‘experience’ was worth it for several reasons:

1. The speaker demoed several ways to share data that I haven’t seen in action since I’ve only been using it to create charts for data sets that are already created for me.

Tableau_Experience

 

2. There were a lot of people there from many different kinds of institutions.  I met others from universities like me – Georgia Tech, Gwinnett College, and so on. But I spent most of the time schmoozing with a lady from Home Depot, and then talked briefly to a man from a company called Norfolk. Everyone was using Tableau for something different – I’m pretty sure I was the only one making an Ebook.

3. I was able to ask about that pesky issue Ryan Cagle and I have been having in trying to get the distance between pane tick marks and labels to be exactly the same in all our charts so they’re uniform within the Ebook. Sorry, Ryan – there is no way to do this. The guy I asked said your idea was the smartest.

Overall, I’m glad I went to the Tableau Experience. And so I leave you with a picture of my spoils, the swanky bathroom in the wetbar, and the view from the 16th floor. Enjoy:

Tableau_Experience-2

Tableau Training

Since the beginning of SIF, I’ve been on the Tobacco Ebook project. I have learned iBooks Author layout and design techniques, which has completely reoriented my thinking about how design works, and increased my attention to detail by… a lot.

This week, I got moved from layout and table building within iBooks Author to recreating charts in a program called Tableau. Tableau can create really clean-looking, beautiful line graphs, bar graphs, and many other types of graphs and charts, like pie charts :). But it’s not easy. First, the data in the spreadsheets the book writers provided us with must be formatted in a way that Tableau ‘likes,’ which is a feat all on its own. If the data isn’t formatted properly, nothing works. Then, the chart has so many formatting options, its enough to make any beginner’s head spin.

Here are two screen shots of graphs I’ve been working with that I can’t seem to get to combine – one line graph and one bar graph:

Tableau line graph

Tableau Bar Graph

The good news is that once I figure out how to make the graphs function the way I like, they are actually pretty easy to make and turn out well. Thankfully, Will knows how to do most of it and is helping me figure out the trickier aspects. Phew!